HPE adds NVIDIA Vera CPU to Private Cloud AI platform

Hewlett Packard Enterprise has unveiled a major expansion of its AI infrastructure portfolio, introducing new hardware and software designed to help businesses deploy and manage autonomous AI agents within secure, on-premises environments. The announcements were made at the HPE Discover 2026 conference in Las Vegas.

The centerpiece is the integration of NVIDIA's new Vera CPU into HPE's Private Cloud AI platform. The Vera processor, purpose-built for agentic computing, will power the new HPE ProLiant Compute DL394 Gen12 server, described as a compute-optimized foundation for autonomous AI and high-performance data processing. HPE executives emphasized that the ARM-based Vera CPU is engineered to eliminate memory bottlenecks common in inference workloads, offering a significant performance gain over traditional x86 architectures for specific tasks. The company stated the Vera-based system will be priced competitively with standard enterprise servers and is expected to be available in the fall.

On the software side, HPE Private Cloud AI will now include the NVIDIA Agent Toolkit. This suite bundles NVIDIA's Nemotron open models, NemoClaw blueprints, and the OpenShell secure runtime,...